| Summary: | 碩士 === 國立中正大學 === 資訊工程研究所 === 103 === With the increase in population, the number of vehicles on the road has increased rapidly, which causes traffic congestion and air pollution.To solve this issue, intelligent transportation systems (ITS) were proposed.
One kind of ITS is traffic signal control systems which adjust signal timing configurations at intersections then the problem of traffic congestion is alleviated.Existing traffic signal control systems adjust signal timing configurations according to historical traffic volumes.However, such an approach is still unable to react dynamically and adaptively to real-time traffic volume.
In order to alleviate traffic congestion dynamically and adaptively, we propose a Cyber-Physical Traffic Control Systems (CPTCS) that integrates the computational elements and physical entities to obtain real-time traffic data and optimize signal timing configurations.Further, we also propose an Adaptive Timing Optimization (ATO) for CPTCS including GA-based signal timing optimization and adaptive adjustment of optimization.The GA-based signal timing optimization tries to optimize signal timing configurations according to the real-time traffic data.To increase scalability of CPTCS, adaptive adjustment of optimization is proposed, which includes adjusting optimization threshold and optimization frequency.
Experiments conducted on optimization show that compared with the fixed timing method, ATO reduces number of waiting vehicles by 34% and incurs reduction in number of waiting vehicles almost double compared with the game theory-based method.Experiments conducted on optimization times show that compared with signal timing optimization only, signal timing optimization with adaptive adjustment of optimization can reduce the optimization times by 21% for a full day traffic.
|
|---|
